2.6.1 Endocrinology, sleep and circadian rhythms

Endogenous circadian rhythms enable organisms to prepare for environmental changes and to temporally modify behavioural and physiological functions. A variation in energy demands appears to be the most important common denominator of these circadian changes, which renders the intimate reciprocal relation of circadian behaviour and endocrine rhythms no surprise. One of the most obvious examples of circadian behaviour is the sleep–wake cycle, closely linked to diurnal variations of locomotor activity, temperature regulation, and water/food intake. Already subtle changes in these circadian cycles may lead to detrimental effects in human biology. Such causative relationship between these changes and adverse biological effects have been obtained not only from mutations characterized in genes responsible for the generation and the integration of circadian rhythms but also from observational studies where circadian rhythmicity was experimentally changed. Life in modern societies tends to increasingly ignore the natural time cues and these environmental insults are increasingly recognized as the underlying mechanism for many pathophysiological changes and a higher susceptibility to disease. Focusing on endocrine-related effects, this chapter will highlight our current understanding of the genetic background of circadian rhythms, their integration with the light–dark cycle and their links to sleep-related changes (1).

Rhythmic circadian behaviour is not restricted to humans but can be detected in the entire animal kingdom, starting with bacteria. These rhythms synchronize biological processes to the day–night (or light–dark) cycle of the natural environment. In humans, it was originally believed that only specialized neurons of the suprachiasmatic nucleus (SCN) are able to induce circadian behaviour but recent detection of rhythmicity in most peripheral organs or cells have challenged this view. The detection of clock and clock-output-genes in these peripheral cells and of rhythmic behaviour when time cues from the SCN are missing suggest a general endogenous pattern. Approximately 5–10% of all peripherally expressed genes show such cyclical behaviour. However, in contrast to the SCN, which integrates time cues from the light circle, peripheral cells are not sensitive to light and are predominantly regulated by hormonal time cues. The time interval between two peaks of an individual rhythm defined as the period is regulated to a 24-h cycle by external time cues, so-called Zeitgeber. When these are missing, individual circadian rhythms may dissociate. These free running rhythms have been detected in experiments with volunteers kept for long time in isolation of all Zeitgebers, especially the two most powerful Zeitgebers, light and food. One of the most striking acute conditions where synchronization of circadian rhythms is temporarily lost is jetlag due to a transmeridian flight. There are multiple other examples for a weakening of the coordination of rhythms. With the recent characterization of the molecular mechanism underlying circadian rhythmicity mutational changes have been described affecting the circadian mechanism in all cells. This may result either in a shorter or a longer than expected underlying rhythm, and external time cues may no longer be able to optimally synchronize circadian rhythmicity. The so-called ‘phase’, which is defined as interval between a fixed event like the beginning of the night and the peak of a given rhythm, is shifted. This can be either shifted to an earlier time, i.e. phase advanced, or may be prolonged or phase delayed (2).

Seminal work in plants, flies, mice, and humans on periodicity genes governing circadian behaviour led to the independent discovery of the first genes involved in circadian behaviour in mammals. Subsequently, the molecular components of the circadian clock were unravelled in much greater detail even though there are still many inconsistencies remaining. A gene named circadian locomoter output cycles kaput or CLOCK and its paralogue neuronal PAS domain protein 2, NPAS2 have been characterized as integral parts of the circadian machinery, which further needs BMAL1 (also known as aryl hydrocarbon receptor nuclear translocator-like; Arntl), period homologue 1 (Per1), Per2, and cryptochrome 1 and 2 (Cry1, Cry2) (Table 2.6.1.1). Under conditions of light, CLOCK (or NPAS2) activates transcription of Per and Cyr proteins by interaction with bmAL1. Per and Cry proteins heterodimerize at high protein levels, and on translocation to the nucleus inhibit their own transcription following interaction with the Clock–Bmal1 complex. Subsequently during the dark period the repressor complex of Per–Cry is degraded, and a new cycle of transcription is activated by Per/Cry. The period of the entire process approximates 24 h. This primary feedback loop is stabilized by a second negative feedback through nuclear hormone receptor Rev-erba. Rev-erba, a direct target of Clock-Bmal1, is a strong inhibitor of Bmal1 transcription. This basic regulation is modulated by a large number of additional factors that change the kinetics of the feedback by altering the stoichiometry of the complexes. During the late afternoon and night Per1 and Per2 proteins are progressively phosphorylated through key kinases such as casein kinase 1δ and ε (CSNK1δ; CSNK1ε). These phosphorylation steps are crucial for the degradation of clock proteins via the proteosomal pathway. Mutants in any members of these regulators may alter the kinetic of the circadian process and result in either short or long periods. Fig. 2.6.1.1 schematically illustrate the process. In addition, more recently the critical importance of circadian changes in histone H3 acetylation and chromatin remodelling for circadian transcription of Clock–Bmal1 target genes has been recognized. It supports an intimate link between the autoregulatory feedback loop and chromatin remodelling (for reviews see Schibler(1), Takahashi et al. (2), Brown et al. (3), and Hussain and Pan (4)).

Intriguing work in a human fibroblast model indicated that these clock genes continue to work as a self-sustained oscillator even outside of the body. It appears that every individual has a given period length which governs its chronotrope. Morning (‘lark’) or evening (‘owl’) types have been characterized in humans who differ by up to 4 h in their optimal cognitive function due to the set-up of the molecular clock (5, 6).

Despite the characterization of clock genes in many mammalian cells the key importance of the hypothalamus for the integration of circadian rhythms is undisputed. Targeted deletion of hypothalamic nuclei including SCN, the ventrolateral preoptic, and dorsomedial nucleus clearly indicates that a normal patterning of sleep–wake cycle, locomotor behaviour, feeding, and the secretion of circadian hormones is no longer observed. Using the same molecular instrumentarium observed in many cells, the SCN appears to be the master regulator of circadian behaviour. This may be based on its ability to respond to light which has been shown for the SCN expression of the clock genes, Per1 and Per2. This sensitivity is detectable only during the night period when levels are low whereas the high levels during the day are not changed by additional light exposure (7).

Studies in nocturnal animals show that a shift in the availability of food has a dominant effect on locomotor activity and the circadian patterns of liver, lung, and heart. This supports the notion that a food-entrainable oscillator exists distinct from the well-characterized light-dependent entrainment of circadian rhythms. Ablation studies indicate that the dorsal medial hypothalamus and not the SCN is critical for food-dependent effects on circadian rhythmicity. A gut–brain communication either via humoral or neural pathways is postulated. Clock genes expressed in the intestinum appear to be independent of the light–dark cycle and processing of food. The circadian rhythmicity observed in triglyceride levels even in the fasting state exemplifies such endogenous diurnal behaviour (4).

The daily pattern of activity versus sleep is the most obvious circadian rhythm in humans. Sleep is controlled by two interacting processes: during the wake phases sleep propensity rises and this dissipates during sleep. This flip-flop mechanism is driven during wakefulness by the monoaminergic systems which inhibit sleep promoting neurons in the ventricular-lateral preoptic regions (VLPO). The firing rate of these sleep promoting, mainly γ-aminobutyric acid (GABA)ergic neurons is high during sleep under the stimulation of adenosine whereas orexinergic and monoaminergic neurons are inhibited. The distribution of neurons that produce orexins (also referred to as hypocretins) is restricted to the perifornical area, the lateral and posterior hypothalamus. Experimental evidence supports a dominant influence of orexin on wake-inducing monoaminergic neurons. Fitting to this pattern, it has been shown in animal models that orexin neurons fire during wake state. They are virtually completely inactive during rapid eye movement (REM) and non-REM sleep (NREMS), a condition associated with atonia. Monoaminergic neurons in turn inhibit sleep-promoting neurons in the VLPO but have also an inhibitory role on orexin neurons, forming a delicate double inhibitory circuit (8). This complex interaction is highlighted by mutations of orexin or orexin 2 receptors in animal models and humans. In both conditions, daytime sleepiness, narcolepsy, and obstructive sleep apnoea may be induced. Narcolepsy with a prevalence of roughly 1 in 2000 is characterized by excessive daytime sleepiness but also a sudden onset of weakness/atonia, cataplexy, fitting to orexin effects on muscle tone. Orexin stimulates locomotor behaviour and energy homoeostasis. Orexin partly stimulates food intake via neuropeptide Y (NPY) but exerts as well an inhibitory action on proopiomelanocortin (POMC) neurons, which will dominantly decrease energy expenditure (8).

In addition to narcolepsy the genetic basis of some other human sleep disorders have been elucidated and linked to circadian alterations in the timing of sleep. The molecular defect in familial advanced sleep phase syndrome (FASPS) has been established as a mutation in the PER2 gene. Patients with this autosomal dominant disease have persistent 3–4-h advanced sleep onset which, however, only leads to clinically apparent problems under a forced sleep–wake schedule. Another clearly genetically based disease, the delayed sleep phase syndrome (DSPS), has been linked to mutations in the CLOCK gene even though the pathophysiology is still not completely unravelled (2, 8, 9).

Again, a close relation to changes in activity/inactivity and food/fasting rhythms, energy homoeostasis, and metabolism is apparent with these mutations. Orexin/ataxin 3 neuron-ablated mice show significantly lower expression of the clock genes Per2, Bmal, and nPas2, along with hypophagia, reduced locomotor activity, and altered energy expenditure. The stimulatory effects of orexin on food anticipation, hunger, locomotor activity, and food intake fit to the close links of the orexin system to multiple metabolic key regulators. Neuronal inputs from the arcuate nucleus through NPY, Agouti, α-melanocyte-stimulating hormone (α-MSH) and also inhibitory inputs from the preoptic GABAergic neurons from leptin and glucose are part of this integrated system. The role of orexin highlights this intimate interplay between sleep–wakefulness, locomotion, and central as well as peripheral metabolic control into a sensitively regulated circadian system. It implies that any disturbance of the sleep–wake cycle, of energy homoeostasis, and of the interfering endocrine regulation may lead to substantial changes in other components of this highly integrated system (8, 9; Fig. 2.6.1.2).

Multiple hormonal systems show pronounced circadian rhythmicity (see Fig. 2.6.1.3 for examples and the link to other circadian rhythms). The physiological relevance of these rhythms is only in part elucidated so far. Multidirectional interactions between sleep, energy homoeostasis, and the endocrine system are currently best characterized, and the following section will thus focus on these interactions. As the capacity of endocrine signals to affect energy homoeostasis are reviewed in more detail in other parts of this book, the following will only review the relevance of energy shifts in relation to circadian behaviour and sleep.

Ghrelin and leptin are among the most important regulators of energy homoeostasis. Both hormones show a significant diurnal variation in lean subjects. Leptin secretion follows a circadian pattern. Both, in lean and obese subjects, a nightly increase of leptin has been shown with peak levels reached at around 02.00 h in the morning hours. Leptin levels decrease thereafter to reach a nadir in the hours between waking and noon. These changes have been viewed as important for the well-known regulatory effects of leptin on energy homoeostasis but also on leptin’s action on other hypothalamic/pituitary hormones (10, 11).

Leptin secretion is linked to another important hormone in appetite control, ghrelin. Ghrelin is released with a marked nocturnal increase with peak fhrelin levels reached after waking in the morning. This diurnal pattern of ghrelin secretion is restricted to lean persons only. An analysis of synchrony between ghrelin and leptin levels neither revealed any copulsatility nor a clear phase shift between the diurnal rhythm found for ghrelin and leptin. It is interesting that the diurnal ghrelin pattern is lost in obese subjects (10).

Ghrelin is known to activate orexin neurons and thus plays an important role in the regulation of food searching behaviour and locomotive activity. Ghrelin also has direct effects on the machinery of circadian behaviour by inducing a phase advance and shifting per2 expression. On the contrary, leptin exerts an inhibitory influence on the firing of orexin neurons and counteracts feeding behaviour by activating POMC. Leptin as well acts directly via periodicity genes as shown in the example of mice deficient in per or Cry in their osteoblasts. At least in this example clear phenotypic changes are found under leptin treatment indicating that leptin acts on osteoblast proliferation via sympathetic nervous pathways and periodicity gene activation (8, 12–14).

With growth hormone-releasing hormone (GHRH) and somatostatin, ghrelin is an important regulator of growth hormone. Growth hormone secretion shows a marked diurnal variation. Using high frequency sampling techniques in several hundreds of volunteers and patients it has been shown that growth hormone is released in secretory pulses, and that these pulses form the basis of a circadian pattern. Modulation of frequency and amplitude of these secretory pulses and their fusion form the circadian rhythm; a common pattern observed in many hormonal systems (13).

Growth hormone in humans is closely linked to sleep. Quantifying the amount of NREMS revealed that GH is robustly associated with the duration and deepness of NREMS but this relation depends on age and gender. It typically develops at about 3 months of age, reaches a peak in adolescence, and progressively decreases after 30 years of age. Despite the fact that there is a marked sex-related difference with a markedly closer relation in males than in females, the decline in slow wave sleep parallels the almost complete decline in nightly growth hormone secretion above the age of 50 in both sexes (14).

These observational studies on a close link between growth hormone and sleep were recently supported by investigations in mutant and transgenic animals. Growth hormone secretion in spontaneous dwarf rats (SDR) is almost completely lost due to a mutation of the GH gene. At variance to expectations NREMS is not reduced in these animals, but rather increased during the rest period. This suggests that growth hormone/insulin-like growth factor 1 (IGF-1) is only indirectly responsible for the reduction in spontaneous NREMS. The suspicion that a major part of this activity is mediated by GHRH- dependent pathways is supported by several mouse models with deletions of the GHRH receptor, such as the lit/lit mice or the dw/dw rats. Growth hormone and IGF-1 productions in both animals are greatly decreased along with significantly reduced spontaneous NREMS. As no GHRH action is expected in these animals chronic growth hormone replacement in these animals allows to dissect GHRH action from growth hormone/IGF-1 responses. Despite successful correction of growth hormone deficiency, growth hormone replacement is unable to stimulate NREMS to normal indicating an important role of GHRH in the regulation of NREMS. This assumption fits to the detection of a circadian and sleep-related variation of GHRH in the hypothalamus. In contrast, REM sleep seems to be directly stimulated by GH secretion (15).

Endogenous cortisol secretion rises sharply between 02.00 and 04.00 h at night with a peak serum concentration approximately 1 h after wakening. Cortisol is secreted in a diurnal pattern which generally reflects the pattern of adrenocorticotropic activity (ACTH). Its high variation between nadir and peak secretions and its high reproducibility allow to use the circadian pattern as a window to evaluate changes in circadian rhythmicity in order to capture pathophysiology. This is exemplified in the diagnosis of Cushing‘s syndrome where the circadian variation is markedly dampened or even abolished. Whereas a morning cortisol may still be within the normal range, typically the midnight cortisol level is increased (16). Similarly, circadian secretion is altered in normal ageing with an increased nightly secretion. This is a mild alteration whereas in depressive illnesses, as well as in chronic alcohol abuse, more pronounced shifts in circadian pattern reminiscent of Cushing’s syndrome are observed. In cortisol deficiency stimulation of the entire corticotropin-releasing hormone (CRH)–ACTH–cortisol axis may be achieved by administration of the 11β-hydroxylase antagonist metyrapone, which blocks the conversion of 11-deoxycortisol to cortisol. The reduced negative feedback inhibition of cortisol on CRH and ACTH secretion can be used diagnostically in partial pituitary insufficiency. It is, however, highly dependent on the circadian timing. Deoxycortisol has been shown to be maximally induced when the drug is applied at 20.00 h. This stimulation was significantly higher than after administration during the morning hours. Similarly, suppression of ACTH secretion by cortisol or synthetic analogues depends on the timing of their administration. Maximal inhibitory effects on ACTH secretion are observed just prior to the endogenous nightly rise in ACTH secretion. These effects have implications for the timing of corticosteroid treatment.

For physiological replacement therapy a slow-release preparation has recently been developed which is able to mimic the nightly cortisol increase (17). Application of the preparation when going to sleep will release peak cortisol levels at the physiological peak secretion time in the early morning hours. It is hoped, but still remains to be proven that these promising preparations will improve the impaired quality of life of patients with Addison’s disease. In pharmacotherapy with glucocorticoids it is evident that the effectiveness depends on the timing. Nightly application improves the effect/dose but side effects are higher as well. Bearing in mind interindividual variations on the sensitivity to glucocorticoids it is currently not worked out which minimal dose is effective at which time of the day (18). In addition, this may vary due to disease specific factors.

Thyrotropin (thyroid-stimulating hormone (TSH)) exhibits a marked circadian rhythm that governs a similar 24-h rhythm of free triiodothyronine (19). Data on the circadian thyroxine rhythm are less clear. There are early observations suggesting a light–dark cycle in total thyroxine but more recent data on free thyroxine could not confirm such dark–night cycle. There are no data on the influence of light on TSH secretion but the important impact of sleep on TSH has been well investigated. Sleep withdrawal induces an acute increase and prolonged release of nocturnal TSH secretion (20; Fig. 2.6.1.4). This is independent of total thyroid hormone levels. In contrast, TSH is almost completely suppressed if the volunteers slept significantly more and deeper in the night following a night of sleep withdrawal. This recovery of hormonal changes in acute total sleep deprivation is observed for other hormonal systems as well and raises the possibility that chronic sleep loss may result in long-term adverse effects via alterations in the circadian rhythmicity. A direct link to energy homoeostasis is currently elusive. Short-term activation of the axis as in acute sleep withdrawal may, however, be linked to an activation of energy stores via the thyroid hormone system, an assumption fitting to data on fasting. Decreased energy availability following a 3-day fast almost completely suppresses circadian TSH release. Effects on the sympathetic nervous system are clearly important in this context but detailed studies are missing to date (21).

Melatonin, which is, exclusively derived from the pineal gland is secreted with a circadian rhythm. Ganglionic photoreceptor cells in the retina integrate information on the light-dark cycle and signal via the retinohypothalamic pathway to the SCN where duration, phase, and amplitude of melatonin hormone production are modulated. Light suppresses melatonin secretion with blue light in the range 460–470 nm having the most pronounced effect. Circulating melatonin dominantly bound to albumin levels shows high interindividual variability which presumably is genetically determined. A circadian rhythm with an increase in the evening hours between 19.00 and 21.00 h, a peak between 02.00and 04.00 h and lowest values during daytime hours seems to be preserved until old age with large interindividual variations (22).

Melatonin exerts its physiological actions through G-protein-coupled specific cell membrane receptors, MT1 and MT2 melatonin receptor. The functions of the subtypes differ and is not restricted to sleep and circadian behaviour where MT1 receptor decreases neuronal firing rates, whereas the MT2 receptor regulates phase shifts. In addition, melatonin is a major regulator of the circadian rhythm of core body temperature in humans. This pattern is linked to sleep. In normal adults, the deepest level of sleep and the lowest core body temperature are reached simultaneously.

Melatonin represents the classical example of a light–dark-driven hormone. It is thought to synchronize circadian rhythms. Investigations in totally blind patients with no recognition of light demonstrate that the coordination of endogenous rhythms is lost. Synchronized endogenous circadian rhythms are important for a normal quality of life. Totally blind persons lose this synchrony and exhibit cyclical sleeplessness associated with daytime sleepiness (23).

Exogenous melatonin affects sleep regulation largely through a phase-resetting mechanism. By its capability to readjust disturbed circadian rhythms to their correct phase position melatonin decreases daytime sleepiness and normalizes sleep quality. Furthermore, other biological rhythms are entrained by melatonin treatment, indicating a general normalizing effect and suggesting a role of melatonin in the coordination of light-dependent effects on the endocrine system. Circadian rhythm sleep disorders, either advance or delayed sleep phase syndrome, have been successfully treated with melatonin (24). A common denominator of these conditions is the loss of coordination between endogenous rhythms. Jet lag induced by a transmeridian flight across several time zones is a well known but transient condition of such loss of entrainment as a mismatch occurs between the endogenous circadian rhythms and the new environmental light–dark cycle (25). Endogenous rhythms shift in the direction of the flight with a phase advance on eastbound flights but a phase delay when flying westwards. Symptoms typically include a disturbed night-time sleep, impaired daytime alertness and performance, irritability, distress and appetite changes, along with other physical symptoms such as disorientation, fatigue, gastrointestinal disturbances and light-headedness. Modification of melatonin secretion has been used to synchronize endogenous rhythms in jetlag and in shift workers. The convincing positive data in the totally blind on a coordinating role of melatonin have recently been paralleled in healthy volunteers treated with a melatonin agonist. The dose-dependent effect on sleep propensity supports an important coordinating function of melatonin on the sleep–wake cycle and on other circadian rhythms. To understand melatonin action in normal physiology, further mechanisms such as the activation of the sympathetic nerve system which suppress melatonin secretion from the pineal gland, may play a crucial role. Light, even dim light at night leads to a suppression of melatonin and feedback on other rhythms. Recent data on melatonin secretion in postmenopausal women highlights this complex relation. Absolute 24-h melatonin secretion is enhanced in depressed postmenopausal women. In addition, timing of melatonin secretion is altered, showing a delayed morning offset of melatonin. This longer melatonin secretion time fits to results in seasonal affective disorders where a dissociation of endogenous rhythms is observed via a shift in the timing of the circadian rhythmicity. These patients also have an increased melatonin secretion during the winter months. In both groups mood disturbance and sleep are improved by bright light therapy (26).

Increased melatonin secretion may further impact on the risk of insulin resistance and diabetes mellitus. Recent data from three independent groups reveal that a polymorphism of the melatonin receptor 1B, which leads to chronic overactivity of the intracellular melatonin dependent signalling pathways, is an independent and strong risk factor for glucose intolerance and type 2 diabetes mellitus. The mechanism behind this is not entirely clarified but inhibition of insulin secretion from pancreatic β cells and a negative impact on incretins seem to cooperate.

Glucose tolerance, which critically depends on the ability of the pancreatic β cell to respond to a given glucose challenge, varies over the day in healthy individuals. It is much lower in the evening than in the morning. There is a further increase in plasma glucose when tested in the middle of the night, suggesting minimal glucose tolerance during sleep. Whereas reduced glucose tolerance in the evening hours is attributed to both, a reduction in insulin sensitivity and a reduced insulin secretory response to glucose, the further deterioration of glucose tolerance during the night depends on sleep related processes to maintain stable glucose levels during the extended overnight fast. During NREMS glucose utilization is lowest; it increases during REM sleep and is highest in the wake period. The underlying multifactorial causes of this circadian change in glucose tolerance are only partly unravelled. Insulin sensitivity decreases in the evening predominantly due to a decreased pancreatic insulin secretory response to glucose. Melatonin-mediated effects may significantly contribute to this regulation. Glucose production and utilization fall in association with sleep during the first half of the night and increase again in the latter part. Insulin-dependent and -independent glucose disposal is reduced during sleep. In parallel, growth hormone secretion is increased with the initiation of slow-wave sleep, cortisol is inhibited, sympathetic nerve activity is decreased, and vagal tone stimulated.

Moderate alteration of night sleep with a reduction to only 4 h/night over a period of 6 nights has been shown experimentally to profoundly affect energy metabolism. It acutely reduces insulin release predominantly via an increased sympathetic outflow and decreases peripheral insulin sensitivity on several levels. Importantly, counteractive hormone release is activated with an augmented nightly growth hormone, TSH, and cortisol secretion, and also with a higher level of cytokines and inflammatory markers (14). It is not surprising that testing for insulin resistance in such a state of sleeplessness revealed a metabolic state well comparable with metabolic syndrome and prediabetes. Similar data have subsequently been obtained in subjects where selective suppression of slow-wave sleep decreased the quality but not the duration of sleep.

The sleep reduction best investigated in obstructive sleep apnoea is further associated with a dysregulation of the neuroendocrine control of appetite. A combined alteration of ghrelin, orexin, and leptin secretion is part of the pathomechanism leading to excessive food intake, decreased energy expenditure and, as recent data indicate, to hypertension (see Fig. 2.6.1.5 for schematic integration of mechanisms). There is further evidence that metabolic changes in the polycystic ovary syndrome are a result of obstructive sleep apnoea. The experimental studies on partial sleep loss and their impact on energy conservation parallel epidemiological findings on the greatly increased risk of obesity and diabetes mellitus in societies. It is tempting to speculate but remains to be proven that sleep curtailment by modern lifestyle changes is a primary force behind the adverse metabolic effects via their impact on diurnal endocrine regulation (27, 28).

In summary, these latter examples clearly demonstrate a powerful circadian regulation of the endocrine/metabolic system interlinked with sleep. Evidence is accumulating that the common curtailment of normal sleep in modern society has important consequences for metabolic and endocrine functions. Data on shift workers who most frequently experience gastrointestinal disturbances support the importance of food-entrained rhythms in addition to the light–dark cycle for the timing of many endocrine rhythms and sleep-associated cycles. Constant violation of these patterns may lead to detrimental effects. The example of treatment with melatonin and melatonin agonists suggests that a better understanding of the pathophysiology of the circadian patterns may help to develop new means to endocrinologically modulate these cycles for the benefit of patients.